Proposed is a continuing evaluation of the Baltimore Needle Exchange Program (NEP), to examine its impact on HIV seroincidence, levels of syringe sharing, health care utilization and referrals to drug treatment. Continued studies of the long-term impact of NEP, and specific components that optimize its effectiveness, are crucial for public health planning and informing policy. An important new aspect of our evaluation focuses on the effectiveness of NEP as a bridge to drug treatment. Based on our experience only one third of NEP attenders requesting referrals actually enter treatment, and 68 percent return to NEP to exchange needles within a year. We recently showed that IDUs who remain in treatment for at least one year are 5 times less likely to return to NEP taking into account levels of ongoing drug use. These data provide a powerful rationale for using a second-tiered intervention at NEP to reinforce the referral process and improve treatment outcomes. The specific aims of this project are: 1) To continue to compare rates of drug-related risk behaviors, HIV seroincidence and health care utilization patterns among participants and non-participants of the Baltimore NEP, and to evaluate innovative components such as i) van vs. pharmacy-based venues, ii) daytime vs. evening NEP services; 2) To determine whether a strengths-based case management (SBCM) intervention significantly improves rates of entry, re-entry and retention in drug treatment among NEP attenders referred to opioid-agonist drug treatment, 3) To determine the cost-effectiveness of a SBCM intervention aimed at improving outcomes achieved through referring NEP participants to opioid-agonist drug treatment programs. To meet Aim 1, comparisons of NEP attenders vs. non-attenders will be conducted using survival analysis and Poisson regression. Descriptive analyses will be used to compare IDUs accessing various types of NEP services. For Aim 2, we will conduct a randomized trial of SBCM vs. standard referrals (i.e., 200 subjects per arm). Follow-up will continue 18 months post-randomization. Logistic regression will be used to identify predictors of entry vs. non-entry, treating subject allocation as a covariate. Survival analysis will be used to compare retention across groups. For Aim 3, we will assess the costs of drug treatment with and without the intervention, and will determine whether the additional cost of SBCM is justified by the economic returns from lower rates of drug use, morbidity, economically motivated crime, and higher levels of employment. In meeting the above specific aims, the proposed study will have an important impact on policies surrounding NEP in the United States and elsewhere.